Ask Slashdot: Ambitious Yet Ethical Software Jobs? 559
First time accepted submitter hwaccaly writes "I'm a mid-career developer with a fair amount of experience working on data-intensive, mathematically ambitious software projects for fun — things like physics and systems simulations, written mostly in CUDA, targeted at Tesla GPUs and small clusters. Ideally, I'd like to get paid for this kind of work, but I've found little call for these skills outside of the financial and defense industries. My conscience won't allow me to accept money from either. The medical/pharmaceutical industries undoubtedly require complex software, but the unavoidable animal testing at the end of the pipeline probably lifts its body count higher even than the defense industry's. And academia pays in degrees, not dollars. So what's left? Do any ethical businesses have a pressing need for high-performance computing, or is it basically a hobbyist niche?"
Ex-Gaming (Score:5, Interesting)
Fifteen years ago I was happy to be in the games industry and saying, "Isn't it nice to have a job for smart technical people that can't possibly be of any use to the military", but now even that's not the case. Plus the industry is wildly volatile and not great or long-term working conditions.
Come work for us (Score:4, Interesting)
We are a very socially conscious company which uses GPU's for video encoding - http://www.elementaltechnologies.com/company/careers/opportunities-at-elemental
If you're into the realm of too ambitious: AI (Score:4, Interesting)
Bare with me: It isn't CYC, but something based on CYC, I forget the actual name. The premise is that you use a 3d simulation to be imagination space for the AI. You need to write an effective physics simulation and database objects into it. If you write an effective 3d imagination space, you could then talk to the 3d imagination space in natural language. The next step is writing vision/laser detection and other senses to read in the real world and simplify it to the imagination space. Once you got something that can turn its environment into something it can think about and do tasks, you have AI. AI isn't some complex and unable to be understood idea where a machine has thoughts like a human, it can be made like a program that just follows orders. Sure once you had AI, you could fake a personality such as by setting coefficients for desiring to do different tasks.
This project would be a lifetime en devour though. I'd be doing it myself if I had enough resources to survive on for the rest of my life. Alas, I need to try and make video games for the short term, so I can have a shot at having it made to do this science work.
I'd aim small to begin with:
3d imagination space, I'd work with as elementary as objects as I could:
Sphere
Block
Rectangle block
Then I would build complex objects out of them. Just this exercise in and of itself could lead to better and bigger things.
Even though it would be many years down the line, the same goes for when you do vision/laser range finding senses to detect the world:
You'd have a really elementary room, like factories. Modern day robots do vision detection, but on a limited number of things to view: Holes to put screws in mainly. So start with just a room with some spheres and blocks in it, and see if the AI can properly observe what is going on. You don't even need a body, just observe what happens in the room.
To me, AI seems very ambitious, but at least there is a plan to do it. Some people can't even grasp that AI is doable. But it is.
A: Write an imagination space that understands natural language.
B: Do vision detection algorithms that map real world objects to imagination space.
C: Have someone build for you a robot that performs any number of functions, slap the AI in, and you're set.
Mind you imagination space and vision detection algorithms might take a man 50 years to do on his own if he is even capable of doing them at all. You'd really think someone like DARPA or something would be working on this and crank it out in 20 years with a crack team of programmers. And hey maybe they are for all we know:P
Fallacious step (Score:3, Interesting)
The universe can only be experienced through a single life, no more, no less, so the destruction of any one life is the destruction of an entire universe of experience. For that reason, the "badness" of that death is infinite.
Taking the two implicit premises as true (infinity of the universe and uniqueness of experience), the "infinity of experience" conclusion is fallacious. The universe may be infinite, but any one person's experience is not necessarily so. In fact, I would tend to think personal experiences are finite and unequal.
Re:About medical... (Score:5, Interesting)
I agree big Pharma is a nasty business
I say a little atheist prayer every night thanking the FSM for the existence of Big Pharma, because without them I'd be in a sanitarium (or whatever they're called now) having almost continuous seizures.
Re:No such thing as an "Ethical" business .... (Score:1, Interesting)
All the work and chemicals used in your pretty garden instead of wirking at a community garden helping to grow fruits and vegetables for less fortunate in your community?
There are ethical dilemmas in most of our actions and choiceswe face daily. Maybe practice mindfulness or maximizing the good you think you're contributing even in a corrupt unethical machine.
It up to you. Ultimately you're the one you face in the mirror and need to be in the good with, not us blathering fools on /.
Ethical Dilemma (Score:5, Interesting)
I faced this very question right at the start of my IT career, in 1968. I had been absolutely against arms manufacture, but was given a chance to move from chemistry/thermodynamics (working in the development of domestic gas burners) to a programming job in aerospace. I have loved aeroplanes since I was 5, an avid SF reader, and going from a "budget" of 30 minutes of mainframe time per week (that was FORTRAN so included compile, test, run) to being 100% programming in technical problems was like being invited to the best party ever. I was going to have to accept a small pay cut, but that didn't matter a bit. Then I realized that every line of code would be used for military aircraft as much or more than for civil projects. It was a long night of the soul, but I decided to take the job. I am so glad I did, not least because I found that most of the military people (real aircrew) were the real anti-war guys. They were the ones most concerned about reducing "collateral damage", and pushing for more accurate delivery of - well - death.
I think we did a good job. Today's wars are still terrible, but compared with conflicts such as WW2 they are actually more controlled, especially when hi-tech systems are used. I am older and wiser now, and doubt that we will ever see an end to war, but I do believe that armed conflict is getting "cleaner", at least when developed countries are involved. If we get more precise systems then we should be able to bring conflicts to a quicker end, with less damage to civilian areas and the environment.
So my advice is to reflect on the outcome of improving technology by better simulation and then decide on each job offer as it comes. This is true whatever area you look at, the arms industry is investing in "non-lethal" systems, the drug companies in simulation and "in vitro" testing, so both of these provide chances for really good jobs in which you can make a positive difference to the world.
I suspect that this might lose me some karma, but I think that gaming is probably the least ethical area (killing things should never be fun, even in a virtual world), and I personally would never work in the financial sector, but then that's the ethical dilemma we all face.
More understanding of your ethics (Score:4, Interesting)
One interesting aspect of your post is the way you have summarized your ethical restrictions and constraints.
Another interesting aspect of your post is the implied view that you feel many jobs are available to you and these jobs are not OK because of your ethical restrictions.
The Slashdot editorial format is very limited but it sounds like you are using ethics as a way to wall yourself off from several classes of employment.
I would say, revisit your ethical ideas. Ethics is more than a process to wall yourself off from the ambiguities and pain of the world. Ethics is a search for truth. Search is a verb. Go to job interviews, find out all about the kind of projects you might work on. Continue your search for truth.
Re:Medical (Score:3, Interesting)
Other humans ALREADY take that risk, in so-called 'clinical trials'. 'Clinical trials' are actually HUMAN experiments - because an astonishing 92% of drugs FAIL human experiments - after passing animal experiments.
Thus proving that animal experiments do not predict human outcomes, and are only done in order to let the pharmaceutical industry off the hook when NINETY TWO PERCENT of their drugs fail in human experiments, and a large number of those that finally get onto the market also FAIL and have to be withdrawn due to side effects.
Animal testing is medical fraud.
Re:Come work for us (Score:2, Interesting)
I sat though a pitch in 2009 from you guys - you sent a very hot blond just gushing about how wonderful your products would be for military airborne video surveillence solutions. Afterwards she wanted to know where the high level brass was and how she could suck up to them as well.
Maybe you should reevaluate your assement of your employer.
Sorry, but wrong (Score:5, Interesting)
You can't simulate with any certainty how a living thing will react in toto to a new drug--if that were possible we'd save ourselves the hundreds of millions currently spent in clinical testing and we'd just run simulations. But there are things you can usefully simulate even if you lack a full understanding of the biological processes involved. For example, it's fairly routine to simulate a drug's pharmacokinetics based on animal data and analogy to other known drugs. This helps us choose doses for clinical trials, it helps us figure out how many patients we need to test in order to produce robust results. Nothing about this is worthless--good sample size estimates minimize unnecessary patient risk and save money.
Re:Medical (Score:4, Interesting)
Re: Ex-Military (Score:4, Interesting)